In my Mercurial repository, I have some build scripts. However, when TeamCity checks out the repo, it doesn't set the execute bit on them, even though it's set in the repository. The build then fails, as it can't run the scripts, as you would expect. How do I make TeamCity respect the execute bit?
I am running TeamCity 9.0 on Ubuntu Server 14.04.
Changing the VCS checkout mode to Agent rather than Server fixed this problem for me.
To change the VCS checkout mode:
On the project configuration settings, go to Version Control Settings
Click on Show advanced options
Change VCS checkout mode from Automatically on server to Automatically on agent
Related
I am new in Teamcity and I created a Build, with a RunScript step.
I changed nothing in the source control configuration of this step.
The goal of the Build step is to run a script nothing else.
When running the build, Teamcity tries to do a source checkout (not configured) and stays blocked there.
What is the correct Teamcity SVC configuration in order that the build works even when SVC is not configured? Is it even possible or I should configure a fake SVC system like GIT to get it running?
With Teamcity, as with TFS, when you want to handle UI Tests, you should install the Test Agents as Interactive and not as Service.
If you Install them as Services, you will not have access to all the functionalities you need for UI testing.
I have working app on OpenShift server. My question is - how to update openshift's git repo of my application, if I make some changes using ssh acsess to openshift? I mean not using all this stuff with pull/push to my local mashine.
If I understand you correctly, you would like to modify source code without using git. I am not sure why you would want that. All that stuff with pull/push gives you a version control flexibility which can save you a lot of time when you screw up one thing. For example, you push brand new UI to production, which turns out to be buggy. With git, you have flexibility to revert back to previous version, and work on different branch to fix the bug on UI.
OpenShift follows conventional app structure. Git for source control, maven for build, jbosseap(for example) for app server, jenkins for continuous integration, etc. So, when you push using git, OpenShift will automatically build using maven, then deploy to the server.
If you would like to disregard all that advantages that OpenShift has to offer, use rhc ssh appname to directly work on the server.
I am using VCS checkout mode to agent side and also have defined the teamcity.hg.agent.path = c:\program files\mercurial\hg.exe in agent.properties file. Still VCS is not picking up this setting and giving me the following error:-
Test connection failed in Dev :: Stocks :: Build and Package. Cannot find mercurial executable at path 'hg'
Also have defined the path in windows environment variables. Mercurial is not installed on server machine. I have read in documentation that if you are using a agent side checkout then not required. Please guide what I am missing here.
thanks
According to the documentation, if you're using server side checkout, then you don't need to install mercurial on your agent.
But when you set up version control as a VCS root, you do need the TeamCity server to be able to talk to the mercurial server so it can do things like detect changes. The JetBrains documentation specifically says that if you're using a Mercurial VCS root:
Mercurial should be installed in the server machine, and, if
agent-side checkout is used, on the agents.
Oure teamcity server (6.5) configured to checkout sources from SVN. For some build proceess cases I need checkout previous successfully builded version(revision). Can teamcity do this? And if can, how to configure checkout?
It sounds like you're looking for TeamCity Snapshot Dependencies, with the "Only use successful builds from suitable ones" option.
You'd end up with two build configurations:
Performs initial builds on commit to SVN
Has a snapshot dependency on #1, so when this build runs (either automatically - via a trigger - or manually), it grabs the same sources as the last successful build of #1.
Both of the build configurations would use the same VCS root.
I am trying to configure a task in Hudson for a VC++ project. I was able to build a project from the file system with MSBuild task. But when I try to configure the task to check out a bazaar repo to do the build, checkout is always failing in authentication. Bazaar passwordless access is setup on the machine and when I use bzr cmd line, checkout is happening without password. Another post suggested that I should have the id_rsa in C:\Documents and Settings\Administrator.hudson - but that also did not help. In Subversion config I saw a way of mentioning username and password. Is there any way to get around this problem.
I assume you have set up the authentication in the windows equivalent of ~/.bazaar/authentication.conf (use bzr version -v to get the correct location).
Is hudson running as the same user as the one you use to connect with the command-line? Because that will impact which authentication.conf it will try to use.
My hudson is using authentication.conf fine but I run it on Ubuntu.
I solved the problem. The authentication.conf is not being considered in windows. I made the repo accessible through http and configured the bazaar with that URL. It was able to download the repo with http protocol without asking for password. One more thing I did was I created a username in hudson, which I matched with a user having access in the bazaar repo which solved another problem which was asking for a user named pwd.